Mohammad B. Uddin1,2,3, Jürgen Kreyling1, Manuel J. Steinbauer1, Carl Beierkuhnlein1
1) Department of Biogeography, University of Bayreuth, D-95447 Bayreuth, Germany
2) Department of Forestry and Environmental Science, Shahjalal University of Science and Technology, Sylhet-3114, Bangladesh
3) corresponding author: e-mail: belal.uddin@uni-bayreuth.de
Abstract
In the face of rapid global changes, improved understanding of the processes and effects of biological invasion is urgently needed. Especially the highly diverse tropical forests are underrepresented in ecological research on invasions. A formalized literature search in the ISI Web of Science (Thomson Reuters) was conducted regarding invasion in the tropical and subtropical forests. Here we reviewed those 55 articles which focus on invasion related to plant species in tropical forests.
Ecosystem or community structure of tropical forest is altered by biological invasions.
Most of the studies supported the two leading hypotheses for the success of invasion;
the disturbance hypothesis, and the empty niche hypothesis. In general the successful invaders have the basic traits of broad environmental tolerances and high dispersability in tropical forests. However, the importance of physical disturbance and allelopathy by the invasive species appear to be unique pattern in tropical forests.
Research on invasion of tropical ecosystems is biased by a large number of studies on islands and wet tropical forests. Thus research is especially needed in mainland and other tropical forests.
Key words: rain forest, biodiversity loss, invasibility, invasiveness, climate change
Introduction
Biological invasions are recognized as a serious threat to biodiversity (Lonsdale 1999;
Mack et al. 2000; Seabloom et al. 2006). As a consequence, biological invasion are a priority issue in nature conservation. However, most knowledge about invasion processes is available for those areas of the world that are monitored since decades, which is not necessarily the same as the hot spots of biodiversity.
In addition, for the understanding of non-native or alien species, native species pools have to be known, which requires precise historical records for species distributions, which is rarely found in the tropics. As a matter of fact, non-native species are not necessarily invasive but may coexist for a long time period with native species after being established either by purpose or not by humans. In temperate regions and especially in Europe, certain dates are fixed for separating phases of introduction (e.g.
the year 1492). However, such distinct temporal lines are hard to be defined for tropical areas, where human activity, migration and trade are ongoing processes that are reaching back even before the last glaciation period. There, it is difficult even to identify species that are definitively not influenced in their distribution by human beings. And finally, the categories and terms that are applied for invasion processes such as “exotic” species are also difficult and can be misleading in a tropical context.
May be these problems in categorizing invasions and invasive species in the tropics are one reason for the small amount of research approaches. This is in a strong discrepancy with the current loss of biodiversity that is mainly going on in the tropics and that is related to a large part also to invasion processes.
Invasive species are non-native species that are performing a strong impact in the recently occupied region. They can disrupt properties, structure, function, and dynamics of entire ecosystems (Mack et al. 2000). Biological invasions thereby affect economic growth and food security (Naylor 2002). Negative impacts of invasions are not limited to natural systems, but are also threatening secondary forests throughout the world (Fine 2002; Lugo and Helmer 2004). Biological invasions may interact with
Manuscript 1
other components of global change and speed up it. For example, invaders respond positively to the increased atmospheric CO2 (Dukes and Mooney 1999; Dukes 2002).
This prevalence of invaders on elevated CO2 impacts ecosystems which ultimately causes global change (Lewis 2006). Climate change and biological invasions are considered as the most important ecological issues in the near future (Ward and Masters 2007). However, there are hints that the effects of climate change and biodiversity loss on invasibility of plant communities are additive without signs of interactive amplifications.
Tropical forests cover approximately 10% of the terrestrial continents, yet, they harbor between one-half and two-thirds of the species in the world (Lewis 2006). A large number of not yet found species (“unknown diversity”) as well as a high variability within taxa (“hidden diversity”) emphasize the value of these unique ecosystems. Yet, tropical forests include a very diverse array of forests differing considerably in abiotic circumstances and disturbance regimes (e.g. tropical moist or rain forests, dry forests, montane forests, savanna forest, mangrove forests etc.) (Lewis 2006).
While in general the understanding of biological invasion has been ever increasing in the last years (Gurevitch et al. 2011), knowledge about biological invasions in the tropics is scarce. Out of 4883 studies obtained for all years up to November 2010 in the ISI Web of Science with the search string: Topic=(invasion or invasive or "exotic species" or "non-native species" or invader*), 539 studies deal with the tropics broadly, and only 273 studies were conducted in the forests.
Patterns of biological invasions in tropical forests differ from other biomes. First, ecosystems have evolved over very long time, presumably causing saturation in number of species unrivaled by other biomes. Second, competition for light is considered to be the main selective pressure in most of the tropical forests and native species are highly specialized in this regard, thereby leaving few options for non-native species. On the other hand, extremely high numbers of species are available within formerly demarcated biogeographical realms (e.g. continents, islands).
Biological exchange may therefore lead to a homogenization, and, thus, to a much
stronger loss in biodiversity than in the well studied circum-polar biomes of the extra-tropical northern hemisphere.
Here, we review the knowledge about the effects of biological invasions in tropical forest ecosystems. In addition, we are categorizing the drivers of tropical invasions based on the invasibility of the ecosystems and invasiveness of invaders. We ask whether classic hypotheses for biological invasions also apply to tropical forests.
Finally, we identify those tropical regions with high priority for future research.
Formalized literature search
A formalized literature search was conducted in the ISI Web of Science (Thomson Reuters) with the search string: Topic= ((invasion or invasive or "exotic species" or
"non-native species" or invader*) and forest* and (tropic* or subtrop*)); timespan = all years; database = SCI-expanded; refined by: subject areas = ecology and document type = article in November 2010. The search yielded 273 articles. Within these 273 articles, 100 articles fulfilled the condition that they strictly focused on biological invasions. Here, we focused on studies stemming from tropical forests, i.e. located in zonobiomes I and II according to Breckle (2002). 55 of these primary research papers focused on invasions by plants in tropical forests while 25 studies have been conducted on animals, mainly invertebrates (19). Therefore, we focus on biological invasions by plants here.
Effects of biological invasions on ecosystem properties
The breakdown of biogeographic barriers is altering ecosystem structure and function globally (Liebhold et al. 1995; Lövei 1997; Dukes and Mooney 1999; Lonsdale 1999;
Naylor 2002). Within our formalized literature search, declining native species richness is reported most frequently (65% of all articles), while alterations of the quantitative species compositions, which is a precondition for species loss, is studied less often (Figure 1). This corresponds well to previous reviews that are not restricted to the tropics (Meiners et al. 2001; Yurkonis et al. 2005).
Invaders are assumed to affect ecosystem processes in the tropics mainly due to shifts in community composition (Parker et al. 1999). Displacement of native species appears to be facilitated by a negative impact of invasions on recruitment and reproduction of native species (38% and 40%, respectively). Seed germination as well
Manuscript 1
as seedling establishment of natives can be negatively altered (i.e. reduced or even non-existent) below the canopy of invaders (Walker and Vitousek 1991).
Reproduction can be affected by the disruption of plant-insect relationships, i.e.
decreased pollinator activity and pollination success (Ghazoul 2004).
Though, even positive effects of invasion on recruitment and growth of native species is reported by one study (Fischer et al. 2009). Cinchona pubescens and Cinchona calisaya, which has been recognized as the worst invaders globally (ISSG 2006), facilitates the regeneration and growth of native species in the Hawaiian archipelago.
There are significantly more species and a higher proportion of endemic species in Cinchona invaded plots.
As a consequence of maladaptation to the use of resources from invasive species by decomposers and herbivores, invaders tend to produce significant amounts of litter.
This can suppress the regeneration of native species (Barua et al. 2001; Islam et al.
2003). The accumulation of litter may also be an effect of special traits of invaders whose success can be supported by allelopathic compounds in some cases (Walker and Vitousek 1991). In addition to potential suppression of regeneration, altered litter quantity and litter quality slows down litter decomposition and nutrient cycling (e.g.
Dunham and Mikheyev 2010).
Figure 1: The effect of plant invasions on ecosystem properties in tropical forests.
Both, the whole ecosystem (a) and single populations (b) can be affected. Note that some effects are mentioned in both categories. (n = 55 studies)
In addition to suppressing recruitment, invaders are reported for their competitive exclusion effect on native species, i.e. by strong competition for resources such as light or water (Mack et al. 2000; Cordell and Sandquist 2008; Friday et al. 2008).
Biological invasions in tropical forests may alter disturbance regimes, for instance by enhancing fire frequency or intensity. It was found that bracken fern and Lantana camara invasions can alter fire regimes in native ecosystems through greater fuel accumulation which affects fire intensity (Schneider and Fernando 2010; Sharma and Raghubanshi 2010). Ultimately, biological invasions in tropical forests can reduce the
Manuscript 1
above ground biomass production in the native stands (Asner et al. 2009; Kueffer et al. 2007). They accelerate C-cycling in the forest soils, and thereby depleting soil C stocks (Litton et al. 2008).
Exotic invaders are reported to experience less herbivory than native species in the invaded ecosystems in the tropics (Dietz et al. 2004). The same was reported from other ecosystems also (Lake and Leishman 2004).
Taken together, plant invasions in tropical forests are affecting various ecosystem properties, e.g. biodiversity, nutrient cycling and disturbance regimes.
Drivers of plant invasions in tropical forests
Drivers, i.e. processes that facilitate biological invasions, can be attributed to two classes (Lonsdale 1999): (1) related to the invaded ecosystem (i.e. invasibility) and (2) related to the invading species (i.e. invasiveness).
Most studies conclude that physical disturbances act as an effective agent to increase the invasibility of tropical forests (Figure 2). Interestingly, positive (Meyer, 1998;
Fine 2002; Baret et al. 2008; Schneider and Fernando 2010) influences of disturbances are reported. That is, disturbances encourage invasion like other ecosystems (Knops et al. 1995; Kotanen et al. 1998; Ross et al. 2002; Buckley et al.
2003; Gelbard and Belnap, 2003). However, negative (Fairfax et al. 2009) influences of disturbances on invasion are reported, although the positive correlations clearly outweigh the negative correlations.
Different anthropogenic disturbances which are very common in tropics such as fire (Schneider and Fernando 2010), logging (Friday et al. 2008), land use changes (i.e.
conversion of forests into agriculture, urbanization, pasture management) (Joshi et al.
2009) facilitate biological invasions.
Natural disturbances, such as wildfires, storm damage, flooding, drought, herbivory or predation, however, can facilitate invasions (Drake 1998; Horvitz and Koop 2001;
Baret et al. 2008; Edward et al. 2009). Vice versa, inherent ecosystem properties such as shading or litter fall constrain invasions by increasing seedling mortality and
decreasing growth of invaders (O’Connor et al. 2000; Garcia-Robledo and Murcia 2005).
Previously empty niches, i.e. the availability of a surplus of resources compared to the use by a native plant community, can explain the invasibility of tropical ecosystems (e.g. Garcia-Robledo and Murcia 2005).
This mechanism can be suspected to be of low importance in saturated tropical ecosystems that have evolved over long time periods. In addition, it is difficult to quantify even if measurement of empty niche space and its impacts on invasion were conducted (e.g. Kueffer et al. 2007; Baret et al. 2008; Cordell and Sandquist 2008;
Friday et al. 2008; Kurten et al. 2008; Schumacher et al. 2009; Asner et al. 2010 etc.).
However, most of the studies from our formalized search just made an interpretation based on their findings that the invasibility of the ecosystem might be enhanced due to the presence of plentiful resources or empty niche. Though, some studies did some measurement of empty niches and showed how it impacts on invasions in tropical forests which were more interesting. For example, Sharma and Raghubanshi (2010) measured the empty niches as the measurement of light intensity and soil moisture in the tropical dry-deciduous forests in India and found that Lantana camara invasion was stronger under conditions of high light and soil moisture availability. Cordeiro et al. (2004) found significantly higher germination rates of an invasive species in large gaps. More empty niches may be available the more isolated a system is, this explains why isolation of ecosystems increase the invasibility of these ecosystems (Drake et al.
1998; Meyer 1998; Bruehl and Eltz 2010).
Manuscript 1
Figure 2: Different drivers characterize the invasibility of ecosystems in tropical systems as indicated by the investigated studies on biological invasions in tropical forests (n = 55 studies). Two drivers only represented in one study are not shown.
Generally, it can be expected that changes in the environmental regime are supporting changes in the invasibility of tropical forests. This leads to the availability of novel niches on one hand and to reduced competitiveness of established species on the other.
Global climate change and biological invasions are threatening the tropical forests through their mutual interactions (Wright 2005; Lewis 2006). For example, elevated CO2 increased the abundance, biomass of invasive species in the tropics (Litton et al.
2008; Raizada et al. 2009). Similarly, increased N-deposition facilitates the success of biological invasions in tropical forests (Asner et al. 2010). The same findings were also reported from other studies (Vitousek et al. 1987; Vitousek and Walker 1989;
Dukes and Mooney 1999; Dukes 2002).
The ongoing fragmentation of habitats in the tropics as a consequence of land use changes is increasing edge effects (non-saturated species assemblages) and these are supporting invasion (Joshi et al. 2009). It was found that different management mechanisms (e.g. root trenching, regular burning etc.) can decrease the invasibility of tropical forests (Fensham and Fairfax 2006; Kueffer et al. 2007).
The control of invasives can be efficient, but it takes a lot of effort to achieve success.
Dominance and biomass of non-native species were significantly decreased through intensive weeding activities in Hawaiian lowland wet forest (Cordell et al. 2009).
Such kinds of approaches are unlikely to be affected in developing countries.
Elevational gradients within landscapes are providing specific opportunities for invasion because they offer a broad range of ecological conditions that apply to a variety of potential invasives (Dietz et al. 2004; Asner et al. 2009). In many cases higher elevated areas (especially mountains) have always been geographically more isolated from comparable ecosystems. They are thus often less saturated than lowland ecosystems and provide more available ecological niches (Steinbauer et al. subm.).
However, depending on the geographical setting, this is not always the case.
Exotic species with broad environmental tolerance and high dispersal rates are the most successful invaders in tropical forests (Figure 3). Mode of dispersal plays an important role for the invasiveness (Rejmanek and Richardson 1996; Lake and Leishman 2004). Exotic invaders are furthermore reported to produce more numerous seeds combined with increased seed longevity than native species (Drake 1998;
Fourqurean et al. 2010).
Manuscript 1
Figure 3: Characteristics that support the invasiveness of a species in tropical systems are indicated by the studies on biological invasions in tropical forests (n = 55 studies).
Some studies indicate that fast vegetative growth rates and vegetative regeneration strongly facilitate rapid spread of invasive species in tropical forests (Lichstein et al.
2004; Rodrigues et al. 2006; Fischer et al. 2009). The attribute vegetative propagation contributes to the success of exotic species invasion in different ecosystems (Henderson 1991; Rejmanek 2000; Lake and Leishman 2004). In addition, litter decomposition and N-fixation are two attributes which can increase invasiveness (Kueffer et al. 2008; Kurten et al. 2008). Highly decomposable litter and N-fixation increase the invasiveness of pioneer invasive species mainly in nutrient-poor soils. In the Hawaii Volcanoes National Park, an introduced small tree, Myrica faya, native to the Canary Islands, alters ecosystem-level properties and ecosystem development in young nitrogen-poor volcanic soils because of its high nitrogen fixing properties, thereby facilitating further invasions (Vitousek et al. 1987; Vitousek and Walker 1989). On the other hand, chemical toxicity of invaders leaf litter induces allelopathic effects on native species and increase invasibility (Davies and Boulton 2009). Some exotic species actively exude allelochemicals which are highly inhibitory to native
plant species in the recipient communities (Hoffmann and Haridasan 2008; Sharma and Raghubanshi 2010). By such means, invasive species modify the habitat through altering the trophic structure, creating effective dispersal barriers for native species etc. (Bruehl and Eltz 2010). Though Fischer et al. (2009) show that habitat modification by invasive Cinchona even facilitate endemic species.
Invasive species are reported to exhibit more extensive root systems and higher specific leaf area than native species, thus enabling superior capture of resources such as light, soil nutrients or soil moisture (Kueffer et al. 2007). Invasive species furthermore benefit from plastic responses to sudden increases in light availability as a result of physical disturbances (Pattison et al. 1998; Davidson et al. 2011).
Generally, plants in the tropics are strongly controlled by highly specialized herbivores (Dietz et al. 2004). Release from such specialized natural pests in the new environment could therefore be a major explanation of invasions success in the tropics, as indicated by facilitated growth, abundance and habitat expansion of invasive species (DeWalt et al. 2004; Garcia-Robledo and Murcia 2005; Fourqurean et al. 2010).
There are some indications that successful invaders in tropical forests are characterized by high genetic plasticity, giving them the potential for rapid evolutionary change (Dawson et al. 2009a; Davidson et al. 2011). The number of studies in tropical forests, however, is not yet sufficient to explore this aspect in detail.
Hypotheses on plant invasions in tropical forests
Several hypotheses (29) have recently been formulated to explain the success of biological invasions (Catford et al. 2009). However, all of these have been postulated based on three major characteristics being influenced by humans; propagule pressure, abiotic and biotic characteristics. Here, we use the more structured and biogeographical classification of such hypotheses based on these characteristics provided by Hierro et al. (2005) for exploring which theories are supported by the case studies from the tropical forests.
Manuscript 1
The result is clear – the disturbance hypothesis and the empty niche hypothesis are well supported in the tropical forests, while other hypotheses are highly controversial with regard to findings (e.g. diversity hypothesis) (Figure 4). However, while the majority of studies report a positive relation between disturbance regime and invasibility, there are cases where e.g. fast-growing introduced species invade nutrient-poor, undisturbed habitats with low resource availability (Schumacher et al.
2009). More than half of the studies that are dealing with the empty niche hypothesis were conducted on islands (both continental and oceanic). Due to the specific characteristics of islands (e.g unsaturated species pool; low competitive ability of natives; lack of invader-specific herbivores and pathogens), island ecosystems are expected to be more invasible (Elton 1958; Lonsdale 1999) and a comparison to mainland systems is difficult.
Figure 4: Formulated hypotheses (according to Hierro et al. 2005) are supported to a different degree by the investigated studies on biological invasions in tropical forests (dark grey bars orientated to the right). Some even mention contrasting results (light grey bars orientating to the left) (n = 55studies).
A certain degree of propagule pressure is necessary for any biological invasion (Drake 1998; Meyer 1998; Colautti et al 2006; Edward et al. 2009). Its relative role as compared to disturbance and empty niches, however, is not emphasized by the studies
carried out up to now. As indicated above, the enemy release hypothesis could be of high relative importance in the tropics due to the highly specialized herbivory (Dietz et al. 2004). However, Dawson et al. (2009b) found no relationship between the enemy release hypothesis and exotic plant species invasiveness in their meta-analysis.
They argued that more-invasive species were planted more than less-invasive species.
They argued that more-invasive species were planted more than less-invasive species.